The print quality of printed images produced by a printing press has typically been checked, during a production run, by the operators of the printing press, either visually or with the use of optical aids. A classification is determined based on the evaluation of the printed images by operators. The printed product is classified into groups of printed matter with predetermined characteristics, such as, for example, based on certain error characteristics that were the object of the evaluation that was performed. All the printed matter that was examined is divided into a first component or class of good quality and a second component or class of lower quality, typically products which are determined not to be usable or saleable, and the printed image of the printed matter to be evaluated is classified as good or as poor, or as containing errors. The evaluation of the quality of printed matter produced by a printing press, by the printing press operating personnel, is subject to considerable fluctuations. This is because the evaluation depends on the judgment of the person in question, and in particular depends on his or her knowledge and experience. As a result, this judgment may be different from person to person.
In the printing industry, camera systems are now increasingly being used for various applications, such as, for example, in inspection systems, in web inspection systems, or in register measurement systems, with these systems typically being arranged on a printing press or in a machine for processing printed matter. These systems perform their functions, for example, “inline,” or during ongoing production of the printed matter to be produced. This results in a considerable challenge for the respective camera system and for an image processing system which is evaluating the camera's image data. Such challenges are often due to the large amount of data which is produced by such a camera system, and due to the rapid process progression in the production of the printed matter. The problem is exacerbated if the printed matter has recognition features which are difficult to identify using spectral photometry and if it is necessary, during the quality control to, for example, make a reliable evaluation of these recognition features in spite of the short time which is allotted for each evaluation due to the usually high speed at which the printed matter is transported. In addition, in security printing, such as, for example, in the production of bank notes, official stamps, or official documents, each individual recognition feature of the printed matter should preferably be subjected to an evaluation. At the same time, for economic reasons alone, there is the requirement, in particular in connection with high-value printed matter such as is used, for example, in the production of bank notes or official stamps, to keep the amount of waste of such printed matter as low as possible, precisely because of their high material and production costs, as long as this low amount of waste is justifiable after taking into account a predetermined level of quality which must be maintained.
In the typical camera systems, as discussed above, a number of electronic image sensors are used for image capture. In particular, color cameras having at least one image sensor comprising, such as, for example, a CCD chip, and whose light-sensitive pixels send an output signal corresponding to the color captured in the observation area, for example, in three separate signal channels, the color channels usually being used for the colors red, green, and blue are often used.
The output signal of an image capture unit, such as, for example the image data of the image captured by the image capture unit, typically needs to be evaluated using an image processing system that is connected to the image capture unit. This evaluation should be accomplished in such a way that an appropriate and balanced evaluation is made of the quality of the printed matter produced by a printing press. In order to evaluate its quality, the printed matter is preferably checked with regard to various criteria.
DE 103 35 147 A1, which was published subsequent to the priority filing date of the subject application, discloses a method for detecting the state of bank notes in which data of at least two different characteristics of the bank notes are evaluated. The data of the at least two different characteristics of each bank note are linked with one another, and the state of the quality of the bank notes is determined from the linked data of the different characteristics. Provision may also be made for an average value to be determined for each of the different characteristics for a number of bank notes in order to determine the state of the quantity of the bank notes for each different characteristic. Alternatively, an average value may be determined for the linked characteristics for a quantity of bank notes in order to determine the overall state of the quantity of bank notes.
DE 103 14 071 B3, which was also published later than the priority filing date of the subject application, relates to a method for qualitative evaluation of a material having at least one recognition feature. A color image is captured, by an electronic image sensor, of at least the one recognition feature. The image sensor directly or indirectly provides a first electrical signal corresponding to the color image. An analysis device, which is connected to the image sensor, evaluates the first electrical signal. A second electrical signal is derived from at least one reference image and is stored in a data storage unit. The second electrical signal has a corresponding target value for the first electrical signal for at least two different characteristics of the reference image. The first signal is compared to at least two of the target values contained in the second electrical signal. At least the color image of the recognition feature is evaluated in the comparison for a color deviation from the reference image. The recognition feature is checked for belonging to a certain class of recognition features or to a certain geometric contour or to a relative arrangement to at least one other recognition feature of the material. Here, the material may be embodied as a bank note or as an official stamp. Each case mentioned above deals with the testing of a material, or with the testing of an individual piece, in which testing the at least one recognition feature of the material in question is checked in testing processes that are different from each other yet which are running independently next to one another with regard to certain criteria.
DE 102 34 086 A1 discloses a method for signal evaluation of an electronic image sensor using pattern recognition of image contents of a test piece. A decision is made regarding the particular class to which a test piece belongs. The method provides for the content of a captured image of the test piece to be evaluated using a classification function that uses methods of fuzzy logic. It is also possible for several classification functions to be linked to one another to form one overarching classification function.
DE 102 34 085 A1 discloses a method for the analysis of color deviations of images captured by an image sensor. The image signal received by the image sensor is analyzed pixel by pixel.
DE 101 32 589 A1 discloses a method for the qualitative evaluation of printed material having at least one recognition feature. An image of the material to be evaluated is captured by one image sensor, and the geometric contour and/or the relative arrangement of several recognition features relative to one another is measured for this image in an analysis device.